Social Media Data Aids Unemployment Rate Estimates

Social media posts about unemployment can predict official jobless claims up to two weeks before government data is released, according to a study. Unemployment can be tough, and people often post about it online. Sam Fraiberger and colleagues developed an artificial intelligence model that identifies unemployment disclosures on social media. Data from 31.5 million Twitter users posting between 2020 and 2022 was used to train a transformer-based classifier called JoblessBERT to detect unemployment-related posts, even those that featured slang or misspellings, such as "I needa job!". The authors used demographic adjustments to account for Twitter's non-representative user base, then forecast US unemployment insurance claims at national, state, and city levels. The model captured nearly three times more unemployment disclosures than previous rule-based approaches while maintaining high precision. The method also reduced forecasting errors by 54.3% compared to industry consensus forecasts. The approach proved particularly valuable during the COVID-19 pandemic, when it detected the massive surge in unemployment claims in March 2020 days before official statistics were released. According to the authors, the methodology demonstrates how AI models combined with social media data can complement traditional economic statistics and provide real-time insights for policymaking, especially during economic crises.